A methodology review synthesizes 32 outcome studies of child and adolescent inpatient psychiatric treatment published since 1975. The principal finding is that conceptual and methodological shortcomings plague the majority of studies, suggesting that findings are less clear than typically believed. Recurring problems include sketchy, if any, data on diagnostic criteria for subject inclusion, lack of control groups, limited information on prior treatment, inadequate description of the hospital setting and active treatment ingredients, infrequent blind analysis, sparse outcome data, and a nonstandardized approach to collecting information. Eight recommendations are offered to assist researchers in conceiving and designing future investigations.